[R-sig-eco] variation partitioning significane test vegan

[Migun Shakya]

Hi all,
A general question on the use of capscale for partition of variance.

I have used capscale to explain the variation in a soil microbial
community into different measured parameters. However, none of the
test (anova.cca) that i ran to check for the significance of
constraints are significant. Does that mean that the variation
explained are not relevant and should not be used to deduce any
conclusions?

Here is what i did:

> FullRDA<-capscale(UNI.UNW.MAT ~A+B+C+D+Latitude+Longitude, comm=OTU,envbac)
> anova.cca(FullRDA)

> EnvRDA<-capscale(UNI.UNW.MAT ~ A+B+Condition( C+D+Latitude+Longitude), comm=OTU,envbac)
> anova.cca(EnvRDA)

> SoilRDA<-capscale(UNI.UNW.MAT ~  C+D+Condition(A+B+Latitude+Longitude), comm=OTU,envbac)
> anova.cca(SoilRDA)

> GeoRDA<-capscale(UNI.UNW.MAT ~ Latitude+Longitude+Condition(A+B+C+D), comm=OTU,envbac)
> anova.cca(GeoRDA)

> GenRDA<-capscale(UNI.UNW.MAT ~ gen$vectors+Condition(BasalArea + BaseTree+ProximalTrees+RiverDistance+ TreeToSample+ Ca+Moist+Mn+Zn+pH1+K+Month+Latitude+Longitude), comm=RHIZOTU,envbac)
> anova.cca(GenRDA)



I hope the question itself is not as much as the answer.

Migee